The The Impact of COVID-19 on FinTech Lending in Indonesia: Evidence From Interrupted Time Series Analysis

Penelitian ini mengukur dampak pandemi COVID-19 terhadap financial technology (FinTech) lending di Indonesia. Menggunakan data bulanan FinTech yang dipublikasikan oleh Otoritas Jasa Keuangan pada periode 2018M02-2021M04, artikel ini mengukur dampak COVID-19 yang dimulai pada tanggal 2 Maret 2020 terhadap FinTech dengan mengadopsi interrupted time series (ITS) experiment. Estimasi memperlihatkan COVID-19 memiliki dampak negatif pada perubahan level (changes in level) FinTech lending di Indonesia, tetapi perubahan tren (changes in trend) adalah positif. Selanjutnya, COVID-19  berdampak negatif dan secara statistik signifikan terhadap perubahan level Rasio Pinjaman Lancar (s.d. 90 hari). Namun, COVID-19  memberikan efek positif dan secara statistik signifikan pada perubahan level Rasio Pinjaman Macet (>90 hari). Temuan ini merekomendasikan bahwa otoritas jasa keuangan secara intensif mendorong berbagai model baru bisnis FinTech yang inovatif post-COVID19 dalam upaya memperluas inklusi keuangan digital dengan menyediakan pembiayaan bagi masyarakat (P2P) yang tidak tersentuh oleh bank.

budget formulation must be consistent with planning: the more synchronous and consistent the formulation, the better the government's performance will be and the more likely development goals are to be achieved. When entering a general election period, economic policy is often used to influence voting, in that the politicians who run the government and are entitled to compete in forthcoming election will behave opportunistically by using the budget they have access to for particular allocations (Klein & Sakurai, 2015;Chortareas, Logothetis, & Papandreou, 2016). In election years, incumbents running for office tend to increase certain spending to send out a "good signal" that reflects their performance (Nordhaus, 1975;Rogoff, 1990). It is therefore important that planning of budget needs is carried out to minimize the potential non-neutrality of stakeholder behavior in the budget process (Lee, 2014).
The absence of a study that explains the relationship between consistency and political budget cycle practices is the motivation for this study, which aims to analyze the relationship between budget planning and trends in political budget cycle practices of local governments in Indonesia. Indonesia regulates simultaneous elections for regional heads in its implementation of regional autonomy policies including development planning and budgeting authority. This research hypothesizes that the consistency index for planning and expenditure as political budget cycle practices can be controlled. This study predicts a relationship between consistency and political cycles with lower budget expenditure.
As the world's largest developing country (by population), Indonesia's political system is a young democracy with characteristics of low ideological division, loyalty to small parties, and the importance of money politics (Gonschorek, Schulze, & Sjahrir, 2018).
Both communities and regional heads have opportunistic attitudes in taking advantage of the election period. Due to the high cost of regional elections, money politics, especially political dowries and vote buying, cause rampant corruption in elections of regional heads.
To reduce or avoid campaign costs, an incumbent who has a budgetary function in local fiscal policy has an incentive to take advantage of certain budget components that will have the effect of attracting voters, such as grants, social assistance, and financial assistance (Sjahrir, Kis-Katos, & Schulze, 2013). In addition to the large discretion available to regional heads in regional election budgets, accountability at the regional level is still weak and democratization has not contributed significantly to reducing waste in regional government spending (Sjahrir, Kis-Katos, & Schulze, 2014). Budgetary order is one of the important solutions that can be built into the formulation of fiscal policy in the election period, regulated in a self-help budget institutional arrangement that involves the community in fiscal agreements (Sanchez, 2007  This study uses government and subsidies) because this type of expenditure is the largest area of discretion for regional heads in regional elections (Sjahrir et al., 2013;Pierskalla & Sacks, 2018 The selection of control variables that reflect economic conditions is based on the regional income variables , , , ) as used in the study by Bonfatti and Forni (2019).
These are chosen because all localgovernment expenditures are highly correlated with the amount of regional income. This approach is used because the realization of local-government national political parties cause a cycle of expenditure in election year budgets motivated by election administration.
The selection of variables based on demographic conditions is as follows: budget will have a large impact and the population variable (( , ) is used influence on the performance of the regional economy as reflected in the gross regional domestic product , (Sjahrir et al., 2013;Wiguna & Khoirunurrofik, 2021).
The control variable that reflects political conditions is the variable status of the incumbent , , as used in the study of Klein and Sakurai (2015) which measures the possibility of fiscal policy being carried out by a first-period regional head who is allowed to run again. The winning incumbent variable ( , ) is applied because the incumbent uses fiscal policy to increase popularity (Eryilmaz & Mercan, 2015).
The more opportunistic the incumbent is prepared to be in order to win, the more they will take advantage of fiscal policy. The political parties variable ( , ) is used in this study because regional heads who are not affiliated to in this study because population size affects the level of demand for government services, which in turn is reflected in the amount of budget expenditure (Klein & Sakurai, 2015); the unemployment rate variable ( , ) is used because a higher rate of unemployment in the community will encourage local governments to issue social assistance to alleviate poverty levels through means such as providing grants. Moreover, we add the the human development index variable The standard deviation of 11.33 means that the maximum increase and decrease in the consistency variable mean is ± 11.33%. The coefficient of variation of the consistency variable is 0.14%, i.e. the data is close to being homogeneous.   after becoming an elected regional head.
The 2017 local election was held by 96 regional governments and the 2018 election by 164 local governments, as presented in Table 3. In terms of regional heads' participation in local elections, several variables are shown in the table below (Table 4), which compares average regional government spending between regional heads who are running for re-election and regional heads who do not run again.  infrastructure functions than regions led by regional heads not participating in the regional head elections. If further analyzed, as shown in Figure 2, the trend of the average regional government spending led by regional heads who run again for the two years before the local election, one year before the local election, the year for the local In addition to the differences in average expenditure, Table 4 shows that the average consistency index value of local governments led by incumbents who ran again was lower, namely 81.74, compared to regions led by regional heads but did not run again, namely 82.95. From Figure 3, it can be seen that in the year the regional head elections were held, the average value of the consistency index of regional governments led by regional heads who ran again decreased sharply to 73.97% compared to regions led by regional heads who did not run again at 80.49 %.    The findings of this study are in line with those of Klein and Sakurai (2015) and Wiguna and Khoirunurrofik (2021)  and subsidies in the two years before the regional elections of 5.79%, one year before regional elections of 21.58%, and in the year of the regional elections of 18.14% for local governments led by a regional head who is running again,  (Sjahrir et al., 2013).  In the education budget, the three interaction variables indicate an increase in the two years before the election of 4.37%, one year before the election of 10.21%, and the year of the election of 4.48% for local governments led by regional heads who run for re-election.
The increase in spending in the education sector is in accordance with studies by Pierskalla and Sacks (2020) which, in analyzing the education sector budget, found that in an election year, voters' interests (Potrafke, 2010).
Improved health services can convince the public of the quality of regional heads who are seen to "side" with the people.
In the infrastructure budget, the three interaction variables indicate an increase in the two years before the election of 11.63%, one year before the election of 13.35%, and the year of the election of 17.44% for local governments led by regional heads who are running again. value one year before the regional head elections in a region led by a regional head who is running for re-election is associated with or has a positive correlation (relationship) with spending on social assistance, grants and subsidies of 2.37% to 2.43% as compared to regions led by regional heads who did not run again.  (Nordhaus, 1975;Rogoff, 1990).
Opportunistic attitudes of this type lead to inconsistencies in the performance of regional expenditure in election years through the use of the allocation composition of expenditure (Drazen & Eslava, 2010), when in fact the main problem faced in regions is limited resources and revenue available to finance activities.

CONCLUSION
There have been many studies discussing political budget cycles at both national and regional levels. This It is reasonable to suggest that a number of limitations might have influenced the results of this study, such as not controlling for the variable characteristics of the individual regional heads, such as their educational background, or for the condition of the Regional Representative Council which of course influences the determination of the budget plan of political parties in the Regional People's Representative Council (DPRD). In addition, the use of a regional head dummy cannot accommodate the possibility of regional heads who have served for two terms supporting replacement candidates for leadership from the same political party.
Further research could add control variables for candidates from the elected regional heads' political parties because of possible relationships to budget allocations during the election period. It would then be advisable to use data over a longer period to produce a comprehensive analysis. and the role of coalition parties in shaping Indonesian local government spending. Asia-Pacific Journal of Regional Science, 1-24.